Phenotypic and molecular characterization of Rhizobium vitis … · Vol. 55, No. 1, April, 2016 43 Rhizobium vitis characterization from Turkish vineyards Figure 1. Turkish provinces

41

Phytopathologia Mediterranea (2016) 55, 1, 41−53

www.fupress.com/pm ISSN (print): 0031-9465Firenze University Press ISSN (online): 1593-2095

DOI: 10.14601/Phytopathol_Mediterr-15776

Corresponding author: D. Canik Orel E-mail: [email protected]

© 2016 Author(s). This is an open access article distributed under the terms of the Creative Commons Attribution License (CC-BY-4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

RESEARCH PAPERS

Phenotypic and molecular characterization of Rhizobium vitis strains from vineyards in TurkeyDIDeM canIk OReL1, aLPeR kaRaGOZ2, RIZa DuRMaZ2,3 and FILIZ eRTunc1

1 Ankara University Faculty of Agriculture Department of Plant Protection, Ankara, Turkey2 Department of Microbiology Reference Laboratories, Turkey Public Health Institute, Molecular Microbiology Research and

Application Laboratory, Ankara, Turkey 3 Department of Clinical Microbiology, Medical Faculty, Yildirim Beyazit University, Ankara, Turkey

Summary. Crown gall-affected grapevine samples were collected during 2009–2010 from major vineyards, lo-cated in different Turkish provinces. One hundred and three bacterial strains were obtained from 88 vineyards and 18 grapevine varieties; they were tumorigenic when inoculated in tobacco, sunflower and Datura stramonium plants and were identified as Rhizobium vitis using biochemical and physiological tests as well as PCR and specific primers. Nineteen R. vitis strains presented a number of anomalous biochemical and physiological characters. PCR and opine-specific primers revealed the presence of octopine/cucumopine-type plasmid in 82 R. vitis strains, nopaline-type plasmids in 18 strains and vitopine-type plasmids in three strains. Clonal relationship of strains was determined using Pulsed Field Gel Electrophoresis following digestion of genomic DNA with the restriction endonuclease PmeI. The greatest genetic diversity was found for the strains from Denizli, Ankara and Nevşehir provinces. Nopaline and vitopine-types of Rhizobium vitis were detected for the first time in Turkey.

Key words: Rhizobium vitis, opine, PCR, PFGE, similarity rate.

IntroductionGrapevine crown gall is an important bacte-

rial disease in viticultural regions worldwide, and is predominantly caused by the validated species Rhizobium vitis (= Agrobacterium vitis) and rarely by Rhizobium radiobacter (= Agrobacterium tumefaciens) and Rhizobium rhizogenes (= Agrobacterium rhizogenes) (Panagopoulos et al., 1978; Süle, 1978; López et al., 2008). It is probable that the taxonomic position of these species will change, because the Rhizobium ge-nus is extremely heterogeneous. In a recent phyloge-netic study, based on multilocus sequence analysis of four housekeeping genes, Mousavi et al. (2015) pro-posed to transter R. vitis to the Allorhizobium genus, as Allorhizobium vitis.

Rhizobium vitis is a Gram-negative soil-borne bac-terium which is specific to Vitis spp. Tumorigenic Rhizobium spp. are unique in carrying a tumor-in-ducing (Ti) plasmid, which bear genes essential for crown gall development. Virulence requires genes lo-cated on different regions of the Ti plasmid, including the transferred DNA (T-DNA) and the virulence (vir) genes. Upon infection, the T-DNA is transferred and expressed in the grapevine genome resulting in host cell proliferation and gall formation. The bacterium is able to utilize specific amino acid derivatives, called opines, as selective nutritional sources in gall tissues (Petit and Tempé, 1995). Opine synthase genes are also carried on the Ti plasmids of the strains. Agro-pine and mannopine are produced by R. radioba-cter, but not by R. vitis, while octopine/cucumopine (O/C) and nopaline can be produced by both spe-cies. Thus far, vitopine is produced only by R. vitis in grapevine. Rhizobium vitis can produce and catabolize three types of opines, namely octopine/cucumopine,

http://creativecommons.org/licenses/by/4.0/

Phytopathologia Mediterranea42

D. Canik Orel et al.

nopaline, and vitopine (Szegedi et al., 1988; Paulus et al., 1989; Burr et al., 1998; Szegedi, 2003).

Previous studies on R. vitis in Turkey were report-ed by Argun et al. (2002) and Kusek (2007). Strains of the bacterium studied by Kusek (2007) in south eastern Anatolia were all of the O/C-type. Argun et al. (2002) reported that the strains from Nevsehir and Ankara provinces differed, based on their T-DNA structures and according to ITS fingerprint analysis, and all strains were identified as the O/C-type. This is the only detailed but province-limited study of Turkish R. vitis strains.

Characterization of Rhizobium spp. may be achieved using several technologies including bio-chemical tests, fatty acid analysis and molecular methods (Moore et al., 2001). Methods based on DNA restriction profiles generated by rare cut endo-nucleases and resolution of fragments from a few ki-lobases to megabase pair in size by Pulsed Field Gel Electrophoresis (PFGE; Sambrook et al., 1989; Arbeit, 1997; Tang et al., 2000) yield particularly sensitive re-striction profiles, useful for typing bacterial strains, including Rhizobium spp. (Aujoulat et al., 2011).

The aim of the present study was to characterize strains of Rhizobium spp. isolated from crown galls on grapevines in the major viticulture areas of Tur-key, using phenotypic tests and molecular analyses, and to determine the genomic relatedness between the strains using PFGE technique.

Materials and methodsSurveys

Extensive surveys were carried out in major Turkish vineyard provinces between July and Sep-tember 2009–2010. One hundred and three crown gall-infected grapevines were sampled from nine different cities in different geographical regions of Turkey (Figure 1), and a total of 103 bacterial isolates were obtained from 18 different grape varieties (Ta-ble 1). Rhizobium radiobacter (C58) and R. vitis (Tm4, AT1, S4) strains were used as positive controls for biochemical and molecular analyses.

Bacterial isolation

Galls collected from infected grapevines were washed under running water, surface disinfected in 10% sodium hypochlorite and rinsed three times with sterile water. Gall tissue samples were each ground

into sterile water with a pestle and mortar, placed in a sterile tube and left for 30 min in suspension so that bacteria could diffuse out of the tissue. The bacterial suspension was then streaked onto a semi-selective Roy and Sasser (RS) medium (Roy and Sasser, 1983) and incubated for 7 d at 28°C. Colonies with smooth margins, red pigmented centres and white transpar-ent halos were selected for further characterization.

Biochemical and physiological tests

All isolates were subjected to the following bio-chemical and physiological tests, described by Ophel and Kerr (1990) and Moore et al. (2001): 3-ketolactose production; acid production from dulcitol, arabitol, melezitose and sucrose; alkali production from L-tartrate and malonic acid; reaction on litmus milk; growth at 35°C and 2% NaCl. Each test was replicat-ed three times. Results were subjected to SPSS (IBM, Version: 22.0) cluster analysis.

Pathogenicity tests

To determine tumorigenicity, all isolates were inoculated on 3–4 week-old seedlings of Nicotiana tabacum var. Samsun, Helianthus annuus and Datura stramonium. Seedling stems were each punctured with a sterile syringe needle, and 5 μL of bacterial suspensions at a concentration of 108 cell mL-1 were applied to the wounds. The inoculations were rep-licated three times for all strains for each test plant species. Inoculated test plants were maintained in a growth chamber (Digitech, P33) for 6 weeks. Isola-tions were made from galls that formed on the plants by plating on RS medium and identification of the pathogen was confirmed by PCR.

Molecular analyses

PCR and specific primers were used to identify the bacterial isolates and to determine their opine types. As template DNA for PCR reactions, bacterial cells lysed according to Abolmaaty et al. (2000) were used. Cells were incubated in 0.1M Tris-HCl buffer (pH 8.0) including 2% Triton X and 2.5 mg mL-1 so-dium azide at 100°C for 10 min, and 3 μL of lysed cells were used as template.

For the identification of the bacterial isolates, PCR and the PGF/PGR primers were used, which amplify the polygalacturonase (pehA) gene and spe-

43Vol. 55, No. 1, April, 2016

Rhizobium vitis characterization from Turkish vineyards

Figure 1. Turkish provinces surveyed for the presence of Rhizobium vitis in vineyards.

Table 1. Strain information and opine types of Rhizobium vitis used in this study.

Isolate Province Variety Opine type

Gaş1, Gaş3 Denizli Cabernet Frank OctopineDM1, DM2, DM3, DM4, DM5, MG3 BoğazkereMG1 ÖküzgözüGoç1 ÇalkarasıDnzl1, Dnzl2, Dnzl3, Dnzl4, Dnzl5, Dnzl6 UnknownGaş2 Cabernet Frank NopalineMG2 ÖküzgözüGaş4, Gaş5 Cabernet Frank VitopineAk1, Ak2 Manisa Alphonse OctopineAk3 RazakiSrh1 AlicanteSrh2, Srh3 SyrahGmm1 UnknownAl1 Alphonse Lavallée NopalineSrh4 SyrahMnd1, Mnd2 İzmir Alphonse Lavallée Octopine

(Continued)



Isolate Province Variety Opine type

Klck1, Klck2, Klck3, Klck4, Klck5, Klck6, Klck7 Ankara Kalecik Karası OctopineAky1, Aky2 Boğazkere Aky3 Cabernet FrankAky4, Aky6, Aky7 EmirAky8, Aky9, Aky10 SyrahBp1, Bp2 UnknownAky11, Aky12 Syrah NopalineÇbk1, Çbk2, Çbk3 Çbk4, Çbk5, Çbk6 UnknownAky5 Emir VitopineKr1, Kv1, Kv2, Kv9, Ürg3, Ürg4, Ürg5 Nevşehir Emir OctopineKv3, Kv5, Kv6 SyrahKr2, Kr3 Pinot NoirKv7 ChardonnayKv8 ÖküzgözüKv10 MerlotÜrg1, Ürg2 NarinceKv4, Syrah NopalineKv11 MerlotÜrg3 EmirDc1, Dc2 Diyarbakır Şıralık OctopineÇer1, Çer2, Çer3, Çng2 ÖküzgözüÇng1 BoğazkereÇng3, Çng4 Boğazkere NopalineMdn1, Mdn2 Elazığ Boğazkere OctopineAlck1, Alck2 TahannebiAlck3 VanniAlck4 Vanni NopalineUzk1, Uzk2, Uzk3, Uzk4, Uzk5, Uzk6, Uzk7, Uzk8 Edirne Syrah OctopineTkr1, Tkr2 Tekirdağ Merlot OctopineC58 (R. radiobacter)S4 (R. vitis)Tm4 (R. vitis)AT1(R. vitis)

VitopineO/CNopaline

Table 1. (Continued).

45Vol. 55, No. 1, April, 2016


cifically R. vitis strains but not other Rhizobium spe-cies (Szegedi and Botka, 2002). The VirD2A/VirD2C primer pair was used to distinguish R. radiobacter-R. vitis based on different annealing temperatures (Bini et al., 2008); opine specific primers were used for de-termining the presence of each opine (Table 2).

PCR amplifications were each conducted in a reaction volume of 25 μL containing 1× PCR buffer (10×, Promega), 2.5 mM MgCl2 (25 mM, Promega), 2.5 mM dNTPs (100 mM, GeneMark), 0.6 mM of each primer, 5% DMSO (Merck), 1.2 U Taq DNA pol-ymerase (GoTaq Flexi, Promega), and 14.5 μL sterile water. The PCR temperature profiles was performed using the following protocol: predenaturation step at 94°C for 4 min, 32 cycles consisting of denaturation at 92ºC for 1 min, annealing at different tempera-tures according to the primers (see Table 2) for 1 min, extension at 72ºC for 90 s, and final extension step at 72ºC for 3 min. The PCR products were visualized following electrophoresis in 1% agarose and staining with ethidium bromide (Syngene).

Pulsed field gel electrophoresis (PFGE)

Bacterial cells were grown in the following liquid medium described by Moore et al. (2001): sucrose 10 g L-1, casein acid hydrolysate 8 g L-1, yeast extract 4 g L-1, K2HPO4 2 g L-1, MgSO4.7H2O 0.3 g L-1. A shaking incubator was used at 28°C for 24 h for bacterial cell growth. Bacterial concentration was spectrophoto-metrically adjusted at OD600 = 1.0. Plug preparation

and lysis of bacterial cells in plugs were performed as described by Durmaz et al. (2007). Briefly, bacterial cells mixed with 10 μL of proteinase K and an equal volume of 1.6% low melting point agarose were pi-petted into plug molds. Solid plugs were then insert-ed in cell lysis buffer containing 10 μL of proteinase K and incubated at 50°C overnight in a water bath. They were then washed twice with distilled water and five times in Tris-EDTA (Tris-HCl, pH: 8.4, 0.5 M EDTA) buffer and stored at 4°C until further use. Six different rare cut restriction endonucleases (RE) were utilized along with different pulse times for the mo-lecular typing of the isolates by PFGE (Table 3). Re-action mix consisted of 20 U enzyme, 10 μL 1× NEB reaction buffer 4, 1U BSA, 87 μL sterile water per iso-late for a total 100 μL volume. PFGE gel results were analysed using Bionumerics software version 6.01 (Applied Maths, Belgium). Dice coefficient and the unweighted pair group method was used to gener-ate a dendrogram based on 1.5% tolerance value and 1% optimization setting. Strains were grouped based on the similarity with a coefficient greater than 85% to clonal relationships.

Results Biochemical and physiological tests

The majority of the strains tested showed bio-chemical and physiological test characteristics con-sistent with those previously reported (Moore et al.,

Table 2. Primer pairs used for differentiating Rhizobium spp. and for determining opine type.

Primer Sequence Amplicon size (bp)

Annealing temperatures

(°C)Reference

VirD2A/VirD2C

5’-ATG CCC GAT CGA GCT CAA GT-3’5’-TCG TCT GGC TGA CTT TCG TCA TAA-3’

224 54°C Haas et al. (1995)

PGF/PGR 5’- GGG GCA GGA TGC GTT TTT GAG-3’5’- GAC GGC ACT GGG GCT AAG GAT-3’

466 56°C Szegedi and Botka (2002)

OCTF/OCTR

5’GAA TAT GAG AAA TCC GTC TCG-3’5’-ACT CAG AGC TCG TGG CCT TG-3’

475 52°C Bini et al. (2008)

NOPF/NOPR

5’-GCA AAC GTA AGT GTT GGA TC-3’5’- CAA GCG AAT ACT CGA GAC G-3’

394 52°C Bini et al. (2008)

VisF/VisR 5’-CCG GCC ACT TCT GCT ATC TGA-3’5’-CCA TTC ACC CGT TGC TGT TAT T-3’

561 55°C Szegedi and Botka (2002)



2001), while 19 strains gave anomalous results (Table 4). The DM1 isolate from the Denizli province was the most divergent strain as it produced atypical results in four tests. Cluster analysis based on bio-chemical and physiological test results revealed that the R. vitis population here studied separated into three clusters (Figure 2). The first cluster consisted of 80 strains, which in turn is divided into two groups, and each group comprised two subgroups. Seventy-two strains belong to the first subgroup, 70 of which showed 100% similarity with the reference strain S4 of R. vitis. In this group, Alck 2 was differentiat-ed from the other strains by alkali production from malonic acid and acid production from sucrose; strain Alck 3 differed from others in acid production from sucrose (Table 4). The second subgroup of the first branch consisted of eight strains. The second cluster consisted of 17 strains, 15 of which showed the same profile. K2 was differentiated by alkali production from malonic acid and K7 by reaction on litmus milk (Table 4). The third cluster consisted of eight strains. DM 1 was the most divergent strain based on alkali production from L-tartrate and malonic acid, reaction on litmus milk and acid production from arabitol (Ta-ble 4). The reference strain of R. radiobacter (C58) was located on this branch and showed a unique profile compared to the other strains.

Pathogenicity tests

All strains were found to be tumorigenic on test plants. Galls were observed, respectively, on sunflower, tobacco and datura plants at 4, 5 and 6 weeks after the inoculation. Koch’s postulates were fulfilled, as R. vitis, identified by colony morphology

Table 3. Rare-cut restriction endonucleases and working conditions used for PFGE (Voltage: 6 V, temperature: 14ºC, enzyme concentration: 20U)

Enzyme Start Pulse (s)

Finish Pulse (s)

Time (h)

AseI10 85 22 5 45 20

DraI 5 45 20

PacI10 85 22 5 45 20

PmeI 5 45 20

SpeI10 85 22 5 45 20

SwaI10 85 22 5 45 20

XbaI 4 10 22

Table 4. Divergent strains of Rhizobium vitis as indicated by biochemical and physiological tests.

Isolate Province

Alca

li pr

oduc

tion

from

L-t

artr

ate

Alca

li pr

oduc

tion

from

mal

onic

aci

d

Acid

pro

duct

ion

from

mel

ezito

se

Acid

pro

duct

ion

from

dul

cito

l

Acid

pro

duct

ion

from

ara

bito

l

Acid

pro

duct

ion

from

sucr

ose

Reac

tion

on L

itmus

milk

Gaş 1 Denizli + + - - + + +Gaş 4 Denizli + + + + - + +Gaş 5 Denizli + + + + + + +DM 1 Denizli - - - - + + -DM 3 Denizli - - - + - + +DM 4 Denizli + + + - - + +MG 1 Denizli - + + - - + -MG 3 Denizli + + + - + + +Kr 1 Nevşehir - + - - - + +Kr 2 Nevşehir - + - - - + +Kr 3 Nevşehir - + - - - + +Kv 1 Nevşehir - + - - + + +Kv 2 Nevşehir + - - - - + +Kv 7 Nevşehir + + - - - + -Çng 4 Diyarbakır - + - - - + +Mdn 2 Elazığ + + - - + + +Alck 1 Elazığ + + - - + + +Alck 2 Elazığ + - - - - - +Alck 3 Elazığ + + - - - - +

In grey are highlithed the anomalous results.

47Vol. 55, No. 1, April, 2016


and PCR, was re-isolated from galls of the inoculated plants.

Molecular analyses

All strains gave a 466-bp product by PCR and the primer pair PGF/PGR, indicating that they belong to R. vitis (Figure 3). In contrast, none of the strains amplified a product with VirD2A/VirD2C primers at 54°C annealing temperature (data not shown).

Opine types of the strains were determined with primers specific for each gene on R. vitis Ti plasmids (Table 2), and all opine type PCR results are shown in Figure 3 representatively. OCTF/OCTR primers amplified a 475-bp product for 82 strains and the ref-erence O/C-type R. vitis Tm4 strain. NOPF/NOPR primers amplified a 394-bp product for 18 of the strains and the reference nopaline-type R. vitis AT1 strain. VisF/VisR primers amplified a 561-bp prod-uct for three strains and from the reference vitopine-type R. vitis S4 strain.

PFGE

Preliminary experiments revealed that restric-tion endonucleases AseI, DraI, PacI and SwaI did not generate any restriction patterns for the strains tested. Among PmeI, SpeI and XbaI, PmeI was found the most effective RE based on the distinguishable restriction patterns (Figure 4). Preliminary test re-sults of SpeI and XbaI showed indistinguishable and unrepeatable patterns on PFGE. Therefore, all strains were digested with PmeI and differences between

Figure 2. SPSS cluster analysis of Rhizobium vitis strains based on phenotypic and physiological test results.

Figure 3. Representative 1% agarose gel of different opine types (M: Thermo scientific 100bp DNA ladder, DM1, Kr3, Dc1 are O/C-type strains, CG415: O/C-type reference strain; Srh4, Aky11, Kv4 are nopaline-type strains, CG49: Nopaline- type reference strain; Gaş4, Gaş5, Aky5 are vito-pine-type strain, S4: Vitopine- type reference strain, SDW: sterile distille water).



the strains were illustrated in a dendrogram accord-ing to their restriction patterns (Figure 5). Due to six strains having indistinguishable PFGE patterns that did not allow a clear evaluation, 97 strains were evaluated. The dendrogram revealed the presence of 19 clusters, 11 unique profiles and 30 genotypes. To-tal genotypes were calculated by “clusters + unique profiles”. A cluster is defined as the phylogenetic group that consists of at least two indistinguishable strains. A unique profile is characterized as unre-lated to other strains and shows an individual pro-file. The dendrogram divided into two main groups. The first group consists of eight clusters and three unique profiles and the similarity rate of the strains in this group is 36.33%. The second group consists of 11 clusters and six unique profiles for which the similarity rate of the strains is 43.34%.

Strains from Denizli province showed the most genotypic heterogeneity with for unique profiles and 13 genotypes among 20 strains. Vitopine-type isolates, Gaş4 and Gaş5, exhibited restriction pat-terns that were different from all other strains. In this region, except unique profiles, another nine strains located under four different clusters. Strains from

Denizli-Denizler (six strains) were the most het-erogenic subgroup within four genotypes. Strains from Ankara were the second most heterogenic subgroup depicted on the dendrogram with seven different genotypes. The similarity rate of the Thra-ce region isolates was 92.34%. Strains from eastern Anatolia region had indistinguishable profiles ex-cept for two strains, Çng 4 from Elazığ and Alck4, from Diyarbakır. These two strains showed unique profiles on the dendrogram. The dendrogram re-vealed that 71 of 97 strains had 100% similarity. R. radiobacter reference strain C58 located separately on the dendrogram and the reference strain was distin-guishable from the other strains by its unique profile.

Discussion Turkey is known as a centre of origin of Vitis vi-

nifera L. and viticulture. Furthermore, vineyards are very important in this country, comprising 468.8 ha of vineyards producing more than 4 million t of grapes per year (Anonymous, 2014). In a previous study of R. vitis characterization in Turkey, Argun et al. (2002)

Figure 4. PFGE patterns of chromosomal DNA restriction fragments resolved in 1.6% pulsed field grade agarose for Rhizo-bium vitis DNA digested with PmeI (Marker: NEB Low Range PFGE Marker, Working conditions: 5sec→45sec, 6V, 14ºC, 20h). The size of the fragments is indicated in kilobase.

49Vol. 55, No. 1, April, 2016


Unknown

UnknownUnknown

UnknownUnknown

UnknownUnknown

Unknown

Unknown

Unknown

UnknownUnknownUnknownUnknown

Unknown

Figure 5. Dendrogram displaying genotypic differences of a Turkish Rhizobium vitis population. Percent similarity was calculated by the Dice similarity of PFGE (PmeI) restriction endonuclease digestion, constructed using UPGMA algorithm (Bionumerics version 6.01 software) based on 1.5% tolerance value and 1% optimization setting.



characterised strains of the pathogen from Nevsehir and Ankara provinces based on their T-DNA struc-tures and according to ITS fingerprint analysis. This is the only detailed but province-limited study for Turkish R. vitis strains. For the study reported here, we aimed to more clearly understand the diversity of R. vitis strains from all of the main Turkish viticul-ture provinces.

Grapevine crown gall was observed in nine of the 13 provinces, that is in the major viticulture areas of Turkey. Bacterial isolates were collected between July and September from young soft cream-coloured galls. All the 103 strains were found to be tumorigenic, based on inoculation of test plants. Rhizobium infec-tion may be initiated at wounds on canes and trunks, caused by freezing temperatures. The majority of the surveyed area, especially central and eastern Anato-lia, are affected by winter and late spring frosts, that may have contributed to development of the disease.

Although it has been reported that VirD2A/VirD2C primers used at an annealing temperature of 54°C gave a characteristic amplicon only with R. radiobacter and the vitopine-type R. vitis strains of Bini et al. (2008), our vitopine-type strains did not produce this amplicon. However, we confirmed that our strains belong to R. vitis, using the specific prim-er pair PGF/PGR (Eastwell et al., 1995; Szegedi and Bottka, 2002) and using other biochemical and physi-ological tests. It is possible that the virD2 sequence of our vitopine strains differs slightly resulting in this lack of amplification. Bini et al. (2008) also reported that results from the VirD2A/VirD2C primer pair did not always give reproducible results at different annealing temperatures.

Although most of the strains conformed to pre-dicted results for R. vitis in biochemical and physi-ological tests, 19 of the 103 strains showed atypical responses. All strains were tumorigenic and ampli-fied the characteristic pehA product with primers PGF/PGR. Eight out of the 19 different strains were L-tartrate negative. Utilization of L-tartrate is a plas-mid-borne feature of R. vitis strains (Ridé et al., 2000; Szegedi et al., 2005), and these strains may therefore not carry the plasmid. Similarly, Kusek et al. (2005) reported that two out of 21 tumorigenic R. vitis strains from the east Mediterranean region of Turkey were negative for alkali production from L-tartrate and malonic acid.

Seventeen of the R. vitis strains analysed were found to be nopaline-types, and three were vito-

pine-types. Similarly, in other studies characterizing Rhizobium strains, it was reported that 60–70% of R. vitis strains were of O/C, 20–30% of nopaline and 5–10% of vitopine-type (Burr and Otten, 1999; Ridé et al., 2000). In the present study, all of the strains were found to amplify at least one opine gene by PCR. Ar-gun (2001) studied the opine types of 50 R. vitis iso-lates from Central Anatolia according to methods of Hooykaas et al. (1979), and found they were the O/C type. It was reported that all the strains carried the same type of Ti plasmid (Argun et al., 2002). Kusek et al. (2005) and Kusek (2007) also reported the O/C type of R. vitis on their study of how plant growth promoting rhizobacteria affect R. vitis in Turkey.

Strains Gas4 and Gas5 from Denizli province and strain Aky5 from Ankara demonstrated five different PFGE banding patterns, and these strains showed unique profiles in the dendrogram. These strains were found to have vitopine-type Ti plasmids. Opine production is another plasmid-borne feature of Rhizobium spp. (Ridé et al., 2000). Schrammeijer et al. (1998) reported that O/C-types of R. vitis have the virF gene on their Ti plasmid, which may be related to the differences we observed in PFGE profiles of nopaline-type and vitopine-type strains. To the best of our knowledge, the current study is the first report of nopaline and vitopine-types of R. vitis in Turkey.

PFGE is a DNA-based typing method that has been employed for studying pathogen epidemiology (Peters, 2009). It has a high segregation capacity and is highly repeatable. Rhizobium vitis strain S4 has two circular chromosomes (Jumas-Bilak et al., 1998; Slater et al., 2009). The larger chromosome (chromosome I) contains an origin of replication that is similar to other chromosomal origins within the Alphaproteobacteria, while chromosome II has a repABC origin of replica-tion typical of the large plasmids within the Rhizoba-cteriaceae. Slater et al. (2009) also reported that strain S4 has five plasmids, rRNA operons and the extensive sets of essential metabolic genes on the second chro-mosome. Therefore, PFGE was chosen to check and compare the whole genome of the strains and find links between the epidemiology of R. vitis in differ-ent parts of Turkey. Schulz et al. (1993) analyzed the genome of R. vitis by PFGE for evolutionary relation-ships of 42 strains. They identified six genomic groups with RE patterns of the six genome types. XbaI, SfiI and SpeI patterns of 42 strains designated six genomic groups. One opine type was found dominant and subgroups were designated as three octopine, three

51Vol. 55, No. 1, April, 2016


vitopine groups. They did not identify any nopa-line subgroup. Our results show that Turkish R. vitis strains consisted of 19 genomic groups, comprising 11 unique profiles according to PmeI restriction pat-terns. Vitopine-type strains and most of the nopaline-type strains appear in separate groups. It may be that as a plasmid-borne feature, opine type of the strains shows different restriction patterns because of their differences on Ti-plasmids. Tanaka et al. (2006) used PFGE to investigate the physical and chromosomal vir genes of K-Ag-1 R. vitis strain which was isolated from kiwifruit. They compared a kiwifruit strain of R. vitis K-Ag-1 with eight R. vitis strains, and reported that three of five Japanese strains were closely related. Two Japanese strains and three from other countries were found different according to PFGE patterns. Similarly, we did not find correlation between the province where strains were taken and genotypic re-latedness of the strains. For example Ankara strains had six different profiles. Some of the strains from Ankara, such as Kalecik strains, are indistinguishable according to restriction patterns, but Ankara-Akyurt strains had three different patterns. Edirne and Deni-zli provinces are located in different regions of Turkey. However DM2, DM4, and DM5 strains from Denizli province appear in the same genomic group with Uzk7 and Uzk8 strains from Edirne province. This report supports our result that there are significant genomic differences between strains isolated from the same region.

The internal transcribed spacer (ITS) region locat-ed between 16S-23S rRNA genes shows variability in size and sequence within bacterial strains. The ITS region has been used for grouping R. vitis strains by Kuzmanović et al. (2014), and they reported five ge-netic groups. Argun et al. (2002) also reported there are differences on the restriction patterns of the ITS region of the O/C-type Turkish R. vitis strains. It shows that Ti plasmid and chromosomal structure can be different and these data support our results on unique and different profiles as expressed in Fig-ure 5. The IGS region of the ribosomal DNA of R. vitis also has different chromosomal and Ti DNA plasmid characteristics between strains (Otten et al., 1996). Those data show that Ti plasmid and chromo-somal structure can be different, and support our re-sults on unique and different profiles as shown in the dendrogram in Figure 5.

Since Turkey is known as a centre of origin of V. vinifera, it is important to understand variability

of the pathogen strains, whether they are from one source and the importance of imported propagation material as sources of the tumorigenic R. vitis. Vine-yards in Turkey still grow local varieties in many ar-eas. Boğazkere, Öküzgözü, Çal karası and Vanni are some of these local and important grape varieties of Turkey. When we examined the grouping of strains on the dendrogram, most strains from local varieties in one region showed identical patterns. However, some strains were different, such as Çng 4-Bogazkere (from Diyarbakır) and Alck 4-Vanni (from Elazığ). These two varieties are grown in eastern Anatolia and both show unique profiles although other strains from the same region have similar patterns. On the other hand, Denizli is another important grape pro-duction area with 44,000 ha of vineyards. This was where we discovered the greatest genetic diversity of R. vitis. Most of the wine varieties in Denizli are imported from abroad which is also likely to impact diversity of R. vitis in this province.

An understanding of the variability of the crown gall-causing bacterial strains is important for several reasons. This will facilitate examination of the future spread of R. vitis in Turkey. This approach can also be used to investigate suspected introductions of the pathogens into the country on plant material or by other means. In addition, divergent strains may also differ in levels of virulence and thus cause differing economic impacts of crown gall. Future research is warranted to determine how the divergent groups of R. vitis that we have identified differ in these and other important characteristics.

This is the first detailed analysis of R. vitis strains using PFGE to identify the distribution and diver-sity of O/C, nopaline and vitopine-type strains in Turkey, from 18 different grapevine varieties and 88 vineyards from main viticulture provinces of Turkey.

AcknowledgementsThe PFGE part of this research was supported by

Scientific Research Project Office of Ankara Univer-sity (12-H-4347001). PFGE experiments were carried out at Ankara University Biotechnology Institute Central Laboratory and GATA Department of Mi-crobiology. The authors thank Dr. Gokhan SOYLE-MEZOGLU for his guidance on surveys, and Dr. Thomas J. BURR from Cornell University-NYSAES for providing reference Rhizobium strains and his helpful advice on the manuscript of this paper.



Literature citedAnonymous, 2014. http://www.tuik.gov.tr/PreTablo.do?alt_

id=1001. Accession 14.09.2014.Abolmaaty A., C. Vu, J. Oliver and R.E. Levin, 2000. Develop-

ment of a new lysis solution for releasing genomic DNA from bacterial cells for DNA amplification by polymerase chain reaction. Microbios 101, 181–189.

Arbeit R.D., 1997. Laboratory procedures for epidemiologic analysis. In: The Staphylococci in Human Disease (K.B. Crossley., G.L. Archer, ed), Churchill Livingstone. New-York, NY, USA, 253–283.

Argun N. 2001. Orta Anadolu baglarında taç uru’na neden olan Agrobacterium vitis’ in bölgesel dagılımı ve bazı biyolojik özel-likleri üzerine arastırmalar. PhD. thesis. Ankara University, Graduate School of Natural and Applied Science, Ankara, Turkey, 84 pp.

Argun N., M.T. Momol, S. Maden, E.A. Momol, C.L. Reid, H. Celek and T.J. Burr, 2002. Characterization of Agrobacteri-um vitis strains isolated from Turkish grape cultivars in the Central Anatolia region. Plant Disease 86, 162–166.

Aujoulat F., E. Jumas-Bilak, A. Masnou, F. Salle, D. Faure, C. Segonds, H. Marchandin, H., and C. Teyssier, 2011. Mul-tilocus sequence-based analysis delineates a clonal popu-lation of Agrobacterium (Rhizobium) radiobacter (Agrobacte-rium tumefaciens) of human origin. Journal of Bacteriology 193, 2608–2618.

Bini F., A. Kuczmog, P. Putnoky, L. Otten, C. Bazzi, T.J. Burr and E. Szegedi, 2008. Novel pathogen-specific primers for the detection of Agrobacterium vitis and Agrobacterium tu-mefaciens. Vitis 47, 1–9.

Burr T.J. and L. Otten, 1999. Crown gall of grape: biology and disease management. Annual Review of Phytopathology 37, 53–80.

Burr T.J., C. Bazzi , S. Süle and L. Otten, 1998. Crown gall of grape, biology of Agrobacterium vitis and the development of disease control strategies. Plant Disease 82, 1288–1297.

Durmaz R., B. Otlu, A. Caliskan and N. Gursoy, 2007. Acine-tobacter baumannii, Escherichia Coli ve Klebsiella Türlerinin Moleküler Tiplendirmesinde Kullanılabilecek Kısa Süreli Pulsed-Field Gel Elektroforez (PFGE) Protokolü. Ankem Dergisi 21, 113–117.

Eastwell K.C., L.G. Willis and T.D. Cavileer, 1995. A rapid and sensitive method to detect Agrobacterium vitis in grapevine cuttings using the polymerase chain reaction. Plant Disease 79, 822–827.

Haas J.H., L.W. Moore, W. Ream and S. Manulis, 1995. Univer-sal PCR primers for detection of phytopathogenic Agro-bacterium strains. Applied and Environmental Micobiology 61, 2879–2884.

Hooykaas P.J.J., C. Roobol and R.A. Schiperoort, 1979. Regula-tion of transfer of Ti plasmids of Agrobacterium tumefaciens. Journal of General Microbiology 100, 99–109.

Jumas-Bilak E., S. Michaux-Charachon, G. Boutg, M. Ramuz and A. Allardet-Servenet 1998. Unconventional genome organization in the alpha subgroup of proteobacteria. Journal of Bacteriology 180, 2749–2755.

Kusek M. 2007. Asmada (Vitis vinifera L.) Ura Neden Olan Agro-bacterium vitis ’in Tanılanması ve Mücadele Olanaklarının

Arastırılması. PhD. thesis. Çukurova University, Graduate School of Natural and Applied Science, Adana, Turkey, 103 pp.

Kusek M., Y. Aysan, F. Sahin ve O. Cinar, 2005. Doğu Akdeniz bölgesinde bağlardan izole edilen Agrobacterium vitis izolatlarının karakterizasyonu. 6th. Viticulture Symposium, 19–23 Eylül 2005, Tekirdağ, Turkey, s. 35.

Kuzmanović N., M. Ivanović, A. Prokić, K. Gašić, N. Zlatković and A. Obradović, 2014. Characterization and phyloge-netic diversity of Agrobacterium vitis from Serbia based on sequence analysis of 16S-23S rRNA inretnal transcripbed spacer (ITS) region. European Journal of Plant Pathology 140, 757–768.

López M.M., R. Palacio-Bielsa, R. Gonzáles-Abolafio, R. San-tiago and R. Penyalver, 2008. Tumorigenic Agrobacterium rhizogenes (biovar 2) isolated from grapevine in Spain. Plant Pathology 57, 367.

Moore L.W., H. Bouzar and T.J. Burr, 2001. Gram-negative bacteria, Agrobacterium. Laboratory Guide for Identification of Plant Pathogenic Bacteria, (N.W. Schaad, J.B. Jones, W. Chun ed.), Third Edition, APS Pres. ST. Paul, Minnesota, 373 pp.

Mousavi S.A, A. Willems, X. Nesme, P. de Lajudie and K. Lindström, 2015. Revised phylogeny of Rhizobacteriaceae: Proposal of the delineation of Pararhizobium gen. nov., and 13 new species combinations. Systematic and Applied Micro-biology 38, 84–90.

Ophel K. and A. Kerr, 1990. Agrobacterium vitis sp. nov. for strains of Agrobacterium Biovar 3 from grapevines. Interna-tional Journal of Systematic Bacteriology 40, 236–241.

Otten L., P. De Rufray, E.A. Momol, T. Momol and T.J. Burr, 1996. Phylogenetic relationships between Agrobacterium vitis isolates and their Ti plasmids. Molecular Plant-Microbe Interactions 9, 782–786.

Panagopoulos C.G., P.G. Psallidas and A.S. Alivizatos, 1978. Studies on biotype 3 of Agrobacterium radiobacter var. tume-faciens. Proceedings of the Fourth International Conference of Plant Pathogenic Bacteria, Angers, France, 221–8.

Paulus F., B. Huss, G. Bonnard, M. Ridé, E. Szegedi, J. Tempé, A. Petit and L. Otten, 1989. Molecular systematics of bio-type III Ti plasmids of Agrobacterium tumefaciens. Molecular Plant-Microbe Interactions 2, 64–74.

Peters T.M. 2009. Pulsed field gel electrophoresis for molecu-lar epidemiology of food pathogens. In: Molecular Epide-miology of Microorganisms, Methods in Molecular Biology 551 (D.A. Caugant, ed.), Humana Press, NY, USA, 59–70.

Petit, A. and J. Tempé, 1995. The function of T-DNA in nature. In: Molecular Form and Function of the Plant Genome, (L. Van Vloten-Doting, G.S. Groot, T.C. Hall, ed.), Plenum Press, New York, NY, USA, 625–636

Ridé M., S. Ridé, A. Petit, C. Bollet, Y. Dessaux and L. Gardan, 2000. Characterization of plasmid borne and chromosome encoded trait of Agrobacterium Biovar 1, 2 and 3 strains from France. Applied and Environmental Microbiology 66, 1818–1825.

Roy, M. and M. Sasser, 1983. A medium selective for Agrobac-terium tumefaciens biotype 3. Phytopathology 73, 810 (Ab-stract).

Sambrook J., E.F. Fritsch and T. Maniatis, 1989. Molecular Clon-ing a Laboratory Manual, Second edition. Cold Spring Har-

53Vol. 55, No. 1, April, 2016


bor Laboratory Press, Cold Spring Harbor, NY, USA.Schrammeijer B., J. Hemelaar and P.J.J. Hooykaas, 1998. The

presence and characterization of virF gene on Agrobacte-rium tumefaciens Ti plasmids. Molecular Plant-Microbe Inter-actions 11, 429–433.

Schulz T.F., C. Bauer, D. Lorenz, R. Plapp and K.W. Eichhorn, 1993. Studies on the evolution of Agrobacterium vitis as based on genomic fingerprinting and IS element analysis. Systematic and Applied Microbiology 16, 322–329.

Slater S.C., B.S. Goldman, B. Goodner, J.C. Setubal, S.K. Far-rand, E.W. Nester, T.J. Burr, L. Banta, A.W. Dickerman, I. Paulsen, L. Otten, G. Suen, R. Welch, N.F. Almeida, F. Ar-nold, O.T. Burton, Z. Du, A. Ewing, E. Godsy, S. Heisel, K.L Houmiel, J. Jhaveri, J. Lu, N.M. Miller, S. Norton, Q. Chen, W. Phoolcharoen, V. Ohlin, D. Ondrusek, N. Pride, S.L. Stricklin, J. Sun C. Wheeler, L. Wilson, H. Zhu and D.W. Wood 2009. Genome sequences of three Agrobacterium biovars help elucidate the evolution of multichromosome genomes in bacteria. Journal of Bacteriology 191, 2501–2511.

Süle S., 1978. Biotypes of Agrobacterium tumefaciens in Hun-gary. Journal of Applied Bacteriology 44, 207–213.

Szegedi E., 2003. Opines in naturally infected grapevine crown gall tumors. Vitis 42, 39–41.

Szegedi E. and S. Bottka 2002. Detection of Agrobacterium vi-tis by Polymerase Chain Reaction grapevine bleeding sap after isolation on a semiselective medium. Vitis 41, 37–42.

Szegedi E., M. Czakó, L. Otten and C.S. Koncz 1988. Opines in crown gall tumours induced by biotype 3 isolates of Agrobacterium tumefaciens. Physiological and Molecular Plant Pathology 32, 237–247.

Szegedi E., S. Bottka, J. Mikulas, L. Otten and S. Sule 2005. Characterization of Agrobacterium tumefaciens strains iso-lated from grapevine. Vitis 44, 49–54.

Tanaka K., H. Urbanczyk, H. Matsui, H. Sawada and K. Su-zuki 2006. Construction of physical map and mapping of chromosomal virulence genes of the Biovar 3 Agrobacte-rium (Agrobacterium vitis) strain K-Ag-1. Genes and Genetic Systems 81, 373–80.

Tang Y.W., M.G. Waddington, D.H. Smith, J.M. Manahan, P.C. Kohner, L.M. Highsmith, H. Li, F.R. Cockerill, R.L. Thomp-son, S.O. Montgomery and D.H. Persing 2000. Compari-son of protein A gene sequencing with pulsed-field gel electrophoresis and epidemiologic data for molecular typ-ing of methicillin resistant Staphylococcus aureus. Journal of Clinic Microbiology 38, 1347–1351.

Accepted for publication: October 16, 2015 Published online: May 14, 2016

Documents

Phenotypic and molecular characterization of Rhizobium vitis … · Vol. 55, No. 1, April, 2016 43 Rhizobium vitis characterization from Turkish vineyards Figure 1. Turkish provinces